SPIRE HeRS/HeLMS Combined SHIM Maps

This repository contains sub-mm wavelength sky images of the combined Herschel Stripe 82 Survey (HerS; [1]) and Herschel Large Mode Survey (HeLMS; [2]) fields as observed by the SPIRE photometer [3] on the Herschel Space Observatory [4] at 250, 350, and 500 microns.  These regions were selected to overlap with the SDSS-Stripe82 [5] and ACT-CMB [6] fields in a contiguous fashion.  The maps cover areas of approximately 360 square degrees at 6, 8.3, and 12 arcsec/pixel.

The data in this repository improve on previous analyses of these data, which were heterogeneously analyzed, served on non-permanent archives, and occasionally had errors discovered by the user community.  The previous analyses include:- HerS survey maps constructed ~2012 with the SHIM map maker, described in [1]- HeLMS maps constructed ~2015 with the SANEPIC map maker, described in [7]- HerS+HeLMS+XMM maps constructed ~2016 with the SHIM map maker, described in [8].- HerS and HeLMS individual maps constructed ~2019 with the CADE map maker, described in [9].While all of these maps are currently circulating in the community, we expect their availability to wane as time goes on.  Further, in many cases the detailed reduction steps and provenance of the served products are unclear.  In this repository we have uploaded vetted and error-free fits files constructed in a consistent way as follows.

To start, we use the Herschel Interactive Processing Environment (HIPE, version 15.0.1) and the most recent calibration files (version 14.3) to reduce the level zero observations to level one data products.  The code and calibration tree are the last released by the Herschel team, and we do not expect improvements or updates to them in the future.  For the products available in this download, our processing consists of the standard scanline calibration pipeline for SPIRE that comes pre-packaged in HIPE (Photometer Large Map Pipeline) with a few changes. The pipeline converts the detector timelines from products in engineering units (satellite pointing and voltages) to science products (sky pointing and flux units) with corrections for cross-talk, signal jumps, glitches, cooler burps, the low pass filter response, and the bolometer time response. We chose to turn off the temperature drift correction as we found that the SHIM map-maker performs better without it and we opt to use the sigma-kappa deglitcher rather than the wavelet deglitcher.  We also do not perform any destriping or baseline subtraction of the scan lines with HIPE.  The HIPE script is available as part of this download for inspection.

We apply this HIPE reduction to the 33 observation IDs listed below (21 from proposal ID OT2_mviero_2 [HerS; 10] and 12 from proposal ID GT2_mviero_1 [HeLMS; 11]), which are retrieved with HIPE from the Herschel Science Archive (HSA; http://archives.esac.esa.int/hsa/whsa/).

Observation IDs

1342247220

1342247993

1342247994

1342247995

1342247996

1342247997

1342247998

1342248000

1342248001

1342248491

1342248492

1342248493

1342248494

1342248495

1342248496

1342248497

1342248498

1342248499

1342248500

1342249103

1342249105

1342234749

1342236232

1342236234

1342236240

1342237550

1342237553

1342237563

1342238251

1342246580

1342246632

1342247216

1342257362

The level one observations are saved as HIPE-format fits files that are input into the map-maker as follows.  Spatial maps of the astrophysical emission are produced using the SHIM map maker [12, with additional processing details provided in 13], which has been found to out-perform other map makers in terms of retaining fidelity on all angular scales [14].  SHIM is a Herschel-SPIRE specific implementation of an iterative baseline removal and detector noise weighting algorithm originally described in [15].  The data are processed scan-wise, i.e. baselines and weights are calculated per one pass of the photometer across a region of sky.  A noise map is created by propagating detector noise as estimated by the variance of the residuals and computing the weighted inverse sum in the map pixels.  This detector noise propagation methodology gives a better measure of noise variations across the map and is more robust when the number of samples per pixel is low and the standard deviation becomes a poor estimate of the statistical error in the measurement.  Also included in the fits files are exposure and flag information.  The images are in units of [Jy/beam], which can be converted to units of surface brightness in [MJy/Sr] by applying the multiplicative conversion factors 90.646, 51.181, and 23.580 at 250, 350, and 500 microns, respectively.  The error maps are reported as the square root of the variance of the statistical noise in each pixel, and do not include the (large) effect of confusion noise in these maps [see eg 16].  Though we do not explicitly provide map transfer function estimates, the transfer function is 1% to angular scales of ~1 degree, and thereafter falls exponentially with a 3dB point of ~6 degrees.

The FITS files provided in this repository are labeled with the SPIRE band names PSW, PMW, and PLW, which correspond to 250, 350, and 500 microns, respectively. Each FITS file contains a header data unit list (HDU, HDUL) with a primary HDU containing only very basic header info and four image HDUs with more extensive header info and the image data. The table below gives some information on the HDUL. 

Ext number
Name
Type
Format
Units

0
PRIMARY
PrimaryHDU
N/A
N/A

1
image
ImageHDU
float64
Jy/Beam

2
error
ImageHDU
float64
Jy/beam

3
exposure
ImageHDU
float64
s

4
mask
ImageHDU
int32
N/A

The 'image' extension contains the science mosaics calibrated in F_nu, the 'error' extension contains the errors corresponding to the science mosaic (and are derived as described above), the 'exposure' extension is the effective number of seconds of integration in each map pixel, and finally the 'mask' extension contains the image mask, which is currently an array of zeros as no masking has been applied to these images. Each image extension also contains information to reference to the world coordinate system.